LinAlgError: singular matrix

[kamiradi]

'LinAlgError: singular matrix' error pops up when trying to call the pairplot() function.

[mwaskom]

Can't help without a reproducible example, sorry.

[bicycle1885]

Hi, this is a (simplified) case I encountered while working on seaborn.pairplot.

data: data.1000.txt

singular.py:

import pandas
import seaborn
data = pandas.read_table("data.1000.txt", index_col="cell")
seaborn.pairplot(data.drop(columns="cluster").iloc[:,0:6], hue="batch")

commands:

~/w/experiments $ python3 --version
Python 3.6.6
~/w/experiments $ pip3 show seaborn
Name: seaborn
Version: 0.9.0
Summary: seaborn: statistical data visualization
Home-page: https://seaborn.pydata.org
Author: Michael Waskom
Author-email: mwaskom@nyu.edu
License: BSD (3-clause)
Location: /Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages
Requires: numpy, scipy, pandas, matplotlib
Required-by:
~/w/experiments $ python3 singular.py
Traceback (most recent call last):
  File "singular.py", line 4, in <module>
    seaborn.pairplot(data.drop(columns="cluster").iloc[:,0:6], hue="batch")
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/seaborn/axisgrid.py", line 2111, in pairplot
    grid.map_diag(kdeplot, **diag_kws)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/seaborn/axisgrid.py", line 1399, in map_diag
    func(data_k, label=label_k, color=color, **kwargs)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/seaborn/distributions.py", line 691, in kdeplot
    cumulative=cumulative, **kwargs)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/seaborn/distributions.py", line 294, in _univariate_kdeplot
    x, y = _scipy_univariate_kde(data, bw, gridsize, cut, clip)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/seaborn/distributions.py", line 366, in _scipy_univariate_kde
    kde = stats.gaussian_kde(data, bw_method=bw)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/scipy/stats/kde.py", line 172, in __init__
    self.set_bandwidth(bw_method=bw_method)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/scipy/stats/kde.py", line 499, in set_bandwidth
    self._compute_covariance()
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/scipy/stats/kde.py", line 510, in _compute_covariance
    self._data_inv_cov = linalg.inv(self._data_covariance)
  File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/scipy/linalg/basic.py", line 975, in inv
    raise LinAlgError("singular matrix")
numpy.linalg.linalg.LinAlgError: singular matrix

[zhoudaxia233]

Hello, I encountered the same situation, do you know how can I make it work without removing the hue parameter?
@bicycle1885 @mwaskom

[mwaskom]

My guess is that it's getting raised when trying to do a KDE on a single observation. You could use a histogram on the diagonal, instead of a kde, which will probably be more robust.

[zhoudaxia233]

My guess is that it's getting raised when trying to do a KDE on a single observation. You could use a histogram on the diagonal, instead of a kde, which will probably be more robust.

Now it works! Thank you very much mwaskom 😄

[100518832]

I have the same error

[elidhu]

Having the same problem, had to install conda to get it to work without getting the linalg error. I'm on a mac too

[cemanughian]

Im having this issue with kdeplot too

[paulbrodersen]

I think the kdeplot fails when any of the variables is integer (or discrete with large bin sizes). The average k-nearest distance is then 0 (for not too large k), which then screws over the kernel width estimation of the KDE.

[emanuelef]

Similar issue but only when using Python 3, Python 2 with same data works fine.

[mjdv]

Same problem. Another easy working example is using the "Eighth-Grade Pupils in the Netherlands" data set as follows.

import pandas as pd
import seaborn as sns

df = pd.read_csv("http://vincentarelbundock.github.io/Rdatasets/csv/MASS/nlschools.csv", index_col=0)
sns.pairplot(df, hue="COMB")

However, it does work if you remove the COMB column from the data to be plotted.

[mwaskom]

Fixed in #1823